The Perils of the Polling Shop

Accusations of bias put Syracuse professor in spotlight, raising questions about when it's appropriate for applied research to serve partisan ends.
July 10, 2008

Many political scientists who study polling also engage in the practice themselves. Several colleges have even made a name for themselves through affiliated public opinion institutes, and some -- like Marist College and Quinnipiac University -- have become nearly synonymous with the independent polls they're popularly attached to, especially during election season.

But what if the professor conducting the surveys is working not for the university, but for political clients?

It's a question that raises issues common to many faculty members who consult on the side but that also involves the intersection of politics and the independent nature of colleges (and the research they produce). While it's perfectly possible, and even likely, that a polling expert could be relied upon to produce both internal polls for candidates of different parties and independent, nonpartisan public polls, even the appearance of bias can tarnish a researcher's reputation -- and, by extension, that of the college.

The issue most recently made headlines last month, when the campaign of a Democratic congressional candidate in upstate New York complained to Syracuse University about a longtime political science professor at the Maxwell School of Citizenship and Public Affairs who had conducted a poll for the Republican opponent using university resources. The poll, for the record, found the race between Republican Dale Sweetland and Democrat Dan Maffei essentially tied (36 percent-37 percent, with 27 percent undecided) to fill a currently Republican seat many Democrats expect to flip by a wider margin.

In fact, the university administration was already aware of and had approved Jeff Stonecash's outside polling activity: For 24 years, he had a standing agreement with his school's dean to make use of office space and hire student workers (at $12.50 an hour) to staff phone banks.

And while Stonecash -- a registered Democrat -- had been working for his Republican client, he has in the past conducted polls for candidates of both parties as well as nonpartisan organizations. He has already met with Syracuse administrators and voluntarily agreed to stop using university resources (which he said he always reimbursed) for outside polling work, although he may continue the consulting work at off-campus sites, Stonecash said.

"I don’t think there’s any doubt that anybody knew I was doing it," he said. "There was nothing secretive about it in the slightest."

Even though Stonecash's activities don't appear to have been a surprise to the administration, it was the appearance of "partisan" activity that likely motivated the mutual agreement to stop using university resources. “I think the general issues are essentially ... the work being done for partisan candidates," said Kevin C. Quinn, Syracuse's vice president for public affairs. "That was the real issue.”

Understandably, universities worry that their names could become associated with a partisan cause. In this case, it wasn't until Stonecash's polling became an issue in a heated political campaign that the reality of his consulting work became clear.

"I do think there’s a legitimate issue about whether or not someone can use the university resources for this," Stonecash said. "Now, having said that, I have no problem with the decision; I will say I find it somewhat selective. A lot of faculty members do consulting using the university resources, so it does make you wonder."

He does, however, take issue with the contention that the poll was by definition biased toward his client.

"I think everybody has this notion that polling is a partisan activity that you twist the questions and you bias the results, and that’s bullshit, it’s just total crap.... In fact, my biggest struggle with candidates is they want me to write biased questions. I won’t do it."

One advantage of being both a scholar and a practitioner of political polling is the ability to meld theory with real-world experience. Stonecash said he takes students to on-campus polling areas, and his book on the subject was based largely on his own consulting practice.

Educational benefits aside, there are also potential concerns with operating a polling shop on the side, said Patrick Murray, the founding director of the Monmouth University Polling Institute, in New Jersey.

"If it is to simply promote research opportunities for the professor and for students, then the issue of whether you take on a partisan client or an advocacy group ... is probably less important as long as you adhere to the academic and research standards of your university ... and whatever field you’re in," he said.

But if a university operates an independent public opinion center while some of its faculty engage in partisan polling, that could be a problem, Murray continued. "[F]or an instance such as mine where a key part of our mission is to hold public officials to account in terms of the issues of the day … we have to be seen as not having a partisan bias, as providing an independent and objective voice of public opinion.… We do do research for hire, but it’s usually for government entities or for nonprofits who are doing internal research. We don’t do any work for candidates because then that would color our credibility when we put out our independent reports [or] when we put out a public poll about an election."

Besides, there's the problem of accuracy, he said. "It's almost a rule of thumb that you always take a partisan polling result with a large grain of salt in terms of you always assume that it’s going to favor their side by a few points, and we’ve seen that when at election time … usually the Democratic pollsters favor the Democrats, the Republican pollsters favor the Republicans, and the truth is always somewhere in the middle."

When a university is public, there's also the institution's neutrality to worry about. In 2006, the producer of public opinion polls at the University of Massachusetts at Lowell came under fire after The Boston Globe disclosed that he was conducting freelance work, both paid and unpaid, for several political candidates.

In that case -- one in which partisan political consulting could have tainted the legitimate, independent work backed by the university -- Murray was called in to do an audit. "There are loads of academics out there who are doing this kind of work on the side, and even involving the students, which is a good learning opportunity for the students," he said.

Regardless of how common the practice is, many colleges don't seem to have explicit policies in place governing outside polling activities, Syracuse included. Steve Manas, a Rutgers University spokesman, said its Eagleton Institute of Politics is a nonpartisan outfit and that none of the affiliated professors had engaged in "outside polling for specific candidates" since at least 1991, when he began working there. Manas pointed to the university's general conflict-of-interest guidelines, which govern professors' consulting activities and set conditions for reporting them to department chairs and deans.

Similarly, a spokesman for the University of Pennsylvania said he was unaware of any specific rules, besides those on conflict of interest, covering polls for specific candidates.


Be the first to know.
Get our free daily newsletter.


Back to Top